Valuation Techniques for Social Cost-Benefit Analysis: - HM Treasury

(1993) and Shizgal (1999) demonstrate that there is a strong positive correlation between wellbeing
ratings and emotions such as smiling and frowning. Research shows that Duchenne smiles
(i.e. a type of smiling that involves a muscle near the eye called orbicularis oculi, pars laterali,
which can distinguish between true and feigned enjoyment) are correlated with subjective wellbeing
(Ekman et al., 1990). Urry et al. (2004) show that reports of life satisfaction are correlated
with activity in the left pre-frontal cortex of the brain, which is the area associated with
sensations of positive emotions and pleasure. Furthermore, well-being is a good predictor of
health, such as heart disease (Sales and House, 1971) and strokes (Huppert, 2006). Cohen et al
(2003) find that people who report higher life satisfaction were less likely to catch a cold and
would recover quicker if they did. Kiecolt-Glaser et al. (2002) find that people with higher life
satisfaction heal more quickly from wounds. Krueger and Schkade (2008) assess the test-retest
reliability of life satisfaction responses. They question the same sample of women two weeks
apart and find that correlation in life satisfaction responses was about r = 0.59, which relates
closely to Kammann and Flett‘s (1983) results. Krueger and Schkade conclude that these levels
of test-retest reliability ‗are probably sufficiently high to yield informative estimates
for……research‘. Finally, it should also be noted that a given person‘s tendency to over or
understate their true well-being scores due to, for example, social desirability reasons could be
seen as an underlying time-invariant personality trait. If this is the case, it would mean that using
fixed effects estimators would control for this effect in life satisfaction regressions (Powdthavee,
2010; Tella and MacCulloch, 2006).
Interpersonal comparability is an important issue for analysts using cross-sectional or panel
datasets on well-being. Sandvik et al (1993) have shown that individuals are able to recognise
and predict the satisfaction of others, suggesting that SWB is observable and comparable
between individuals. van Praag (1991) and Stutzer and Frey (2003) show that different
individuals easily translate verbal labels, e.g. very good, into roughly the same numerical values.
Furthermore, Kahneman (2000) suggests that there is considerable interpersonal convergence in
the ranking of pleasure and pain.
4.2.3 Setting up the empirical study
All three valuation methods rely on collecting a sample of data from a chosen population in
order to derive valuation estimates. Stated preference applications rely on the construction of
original questionnaires which are then distributed to members of a target population. Revealed
preference and life satisfaction applications typically use secondary data or a combination of
both primary and secondary data.
Collecting primary data has the advantage of allowing value estimates to be based on
specifically defined populations. For example, the direct users of a good (e.g. people with a
licence to fish on a lake) or individuals within a specifically defined geographical area. It can also
lead to bias reduction in certain areas. For example, as discussed in 4.2.2 above, Bertrand and
Mullainathan (2001) argue that responses to subjective well-being questions are vulnerable to
ordering effects. Dolan and Metcalf, (2008) note that this is a problem because many secondary
surveys which ask life satisfaction questions, such as the British Household Panel Survey (BHPS),
situate the (global) life satisfaction question in the middle of the survey.
However, original data collection is susceptible to problems that are potentially less prevalent in
large national surveys, such as non-response bias, interviewer bias, and information bias. Nonresponse
bias can occur when individuals who respond to a survey systematically differ from
non-respondents (Champ, 2003). In contingent valuation surveys, for example, individuals with
particularly strong feelings toward the good in question may be more willing to commit time to
the questionnaire. Interviewer bias occurs when surveys are administered via telephone or faceto-face
and the presence of the interviewer influences responses. This effect can be avoided with
27